ORIGINAL ARTICLE. Primary Stapes Surgery in Patients With Otosclerosis

Transcription

1 ONLINE FIRST ORIGINAL ARTICLE Primary Stapes Surgery in Patients With Otosclerosis Prediction of Postoperative Outcome Arnold J. N. Bittermann, MD; Maroeska M. Rovers, PhD; Rinze A. Tange, MD, PhD; Robert Vincent, MD; Wouter A. Dreschler, PhD; Wilko Grolman, MD, PhD Objectives: To evaluate the audiometric results of primary stapes surgery in patients with otosclerosis and to determine predictors of a postoperative air-bone gap (ABG) of 10 db or less and a postoperative gain in air conduction (AC) exceeding 20 db. Design: Retrospective cohort study. Setting: Tertiary referral center in Utrecht, the Netherlands. Patients: Nine hundred thirty-nine patients with otosclerosis who underwent primary stapes surgery between January 1, 1982, and March 1, Intervention: Primary stapes surgery. Main Outcome Measures: Preoperative and postoperative audiometric results were compared. Logistic regression analyses were performed to evaluate which factors (ie, sex, age at surgery, bilateral otosclerosis, and preoperative 4-frequency [0.5, 1, 2, and 4 khz] ABG, AC, or bone conduction) independently contributed to the prediction of a postoperative ABG of 10 db or less and a postoperative gain in AC exceeding 20 db. Results: A total of 72.1% of patients had a postoperative ABG of 10 db or less, and 93.8% of patients had a postoperative ABG of 20 db or less. Age at surgery and preoperative ABG and AC were independent prognostic determinants. A patient older than 40 years with a preoperative ABG of 30 db or less has a 77.6% chance of achieving a postoperative ABG of 10 db or less. A patient with a preoperative AC exceeding 50 db and a preoperative ABG exceeding 30 db has an 86.2% chance of achieving a postoperative gain in AC exceeding 20 db. Conclusions: Following primary stapes surgery, a postoperative ABG of 10 db or less and a postoperative gain in AC exceeding 20 db may be predicted with accuracies of 62.1% and 80.1%, respectively. Clinicians can use this information to inform patients more explicitly about expected postoperative audiometric results. Arch Otolaryngol Head Neck Surg. 2011;137(8): Published online July 18, doi: /archoto Author Affiliations: Department of Otorhinolaryngology Head and Neck Surgery (Drs Bittermann, Rovers, Tange, and Grolman), Rudolf Magnus Institute of Neuroscience (Drs Bittermann and Grolman), and Julius Center for Health Sciences and Primary Care (Dr Rovers), University Medical Center, Utrecht, and Clinical and Experimental Audiology, Ears Nose Throat Department, Academic Medical Center, Amsterdam (Dr Dreschler), the Netherlands; and Jean Causse Ear Clinic, Traverse de Bézier, Colombiers, France (Dr Vincent). OTOSCLEROSIS IS CHARACterized by disordered bone remodeling in the region of the otic capsule, located primarily between the cochlea and the vestibule and just anterior to the footplate of the stapes. 1,2 In 2001, Declau et al 3 published their findings in an unselected series of temporal bones and calculated that clinical otosclerosis has a prevalence of 0.3% to 0.4% among the white ethnic population. Absolute numbers in 2009 in the United States and the Netherlands can be extrapolated to and , respectively. It is important to distinguish histologically demonstrated otosclerosis (without symptoms) from clinical otosclerosis (with symptoms like hearing loss and vertigo). Histologically demonstrated otosclerosis is about 10 times more common than clinical otosclerosis. 1 Stapes surgery is a treatment option for hearing loss resulting from otosclerosis and has proved to be an effective and safe intervention. 4-8 However, postoperative results show large variability. Among large series of stapedotomies, reported airbone gap (ABG) closure (closed to 10 db) varied from 94% (n=2368) to 75% (n=861). 6,7 Patient characteristics, surgical experience, and intraoperative findings may be considered potential prognostic factors affecting postoperative audiometric results. Studies 4-7,9-11 have described the effectiveness of stapes surgery. Two of these reported 10 potential prognostic factors, but only 1 factor was investigated at a time: Ueda et al 10 found that a small preoperative ABG resulted in better postoperative ABG closure at frequencies under 1 khz, while Gerard et al 11 identified no variables that were prognostic among the factors that they studied. 780

2 Given the variability among patients and in the cause, presentation, and treatment of diseases and other health states, a single predictor or variable rarely gives an adequate estimate of prognosis. Physicians implicitly or explicitly use multiple predictors to estimate a patient s prognosis. Therefore, prognostic studies need to use a multivariate approach in design and analysis to determine important predictors of the studied outcomes, to give outcome probabilities for different combinations of predictors, or to provide tools to estimate such probabilities. These enable care physicians to use combinations of predictor values to estimate an absolute risk or probability that an outcome will occur in an individual. 12 A postoperative ABG of 10 db or less could be considered the primary outcome measurement in the stapes surgery literature 6,7,9 and should be included in a prognostic model evaluating this procedure. However, it is also important to include the postoperative gain in air conduction(ac) because it reflects the degree of success in restoration of middle ear hearing transmission function. 5 A reduction in the ABG could reflect improved AC, deteriorated bone conduction (BC), or a combination of the change in AC and BC. As potential predictors affecting postoperative ABG and postoperative gain in AC, we selected sex, age at surgery, bilateral otosclerosis, and preoperative 4-frequency (0.5, 1, 2, and 4 khz) ABG, AC, and BC. The selection was based on data in the literature, 10,13 clinical expertise, and available information from routine clinical practice. Therefore, we performed a multivariate prognostic study among patients with otosclerosis who underwent primary stapes surgery. We determined predictors of 2 outcome measurements, a postoperative ABG of 10 db or less and a postoperative gain in AC exceeding 20 db. METHODS PATIENTS We performed a retrospective cohort study among 939 patients. These patients underwent primary stapes surgery at a specialized tertiary referral center in Utrecht, the Netherlands, between January 1, 1982, and March 1, SURGICAL PROCEDURE An endaural procedure was performed in all the patients. A stapedectomy was used in 2.7% of cases; a small fenestra stapedotomy was used in the others. The fenestration was performed in most patients using the micropick technique described by Marquet. 14 Potassium titanyl phosphate laser use was introduced at our center in The surgical procedures were performed by 2 of us (R.A.T. and W.G.). Various prostheses were used (Table 1). AUDIOMETRIC ASSESSMENT Audiometric evaluation included preoperative and postoperative ABG, AC, and BC thresholds. Postoperative follow-up data were collected 3 months after surgery, which is the first regular follow-up visit at our institution. We used a 4-frequency puretone mean for ABG, AC, and BC thresholds. Audiometry results were reported according to American Academy of Otolaryngology Head and Neck Surgery Foundation guidelines, 15 except for thresholds at 3 khz, which were substituted in all cases with those at 4 khz. In our series, we focused on Table 1. Baseline Characteristics of the Study Cohort Entire Cohort (N=939) 0.5, 1, 2, and 4 khz because of the availability of these data. It has been shown that the percentage of patients with a postoperative ABG of 10 db or less will increase using the mean thresholds of 0.5, 1, 2, and 3 khz instead of 0.5, 1, 2, and 4 khz. 6 OUTCOME PREDICTORS On the basis of data in the literature, clinical expertise, and available information from routine clinical practice, the following baseline candidate predictors were selected: sex, age at surgery, bilateral otosclerosis, and preoperative ABG, AC, and BC. We aim to use the prognostic model during preclinical assessment; therefore, we only included predictors that were available at the time of diagnosis. 12 STATISTICAL ANALYSIS Patients Without Missing Values (n=666) Characteristic Age at surgery, mean (SD) 41.4 (11.5) [range], y [9-74] Sex, % Female Male Ear operated on, % Left Right (11.2) [10-72] Potassium titanyl phosphate laser surgery, % Surgeon, % R.A.T W.G Bilateral otosclerosis, % Piston type, % Titanium, mm à Wengen, a mm Causse, b length manually fitted Shea Teflon c cup or loop, length manually fitted Gold, mm Other a Manufactured by Atos Medical, Hörby, Sweden. b Manufactured by Medtronic, Minneapolis, Minnesota. c Manufactured by Richards Manufacturing Company, Memphis, Tennessee. To evaluate the effect of primary stapes surgery in patients with otosclerosis, preoperative and postoperative audiometric results were compared. Paired-samples t test was used. The association between each prognostic factor and the presence of a postoperative ABG of 10 db or less or a postoperative gain in AC exceeding 20 db was examined using univariate logistic regression analyses. Predictors that were associated with the outcome in univariate analyses (P.10) were included in multivariate logistic regression analyses. The model was reduced by exclusion of predictors with P.05. The predictive ac- 781

3 Table 2. Preoperative and Postoperative Audiometric Results Among 666 Patients Without Missing Values Pure-Tone, Mean (SD), db Pure-Tone Improvement Variable Before Surgery 3 mo After Surgery Bone Conduction or Deterioration (95% Confidence Interval), db P Value 0.5 khz 15.9 (10.4) 16.9 (11.6) 1.00 ( 1.79 to 0.27) khz 18.5 (10.5) 17.6 (12.6) 0.90 (0.06 to 1.68).04 2 khz 29.7 (12.2) 25.0 (14.2) 4.70 (3.95 to 5.56) khz 25.8 (16.1) 28.1 (18.3) 2.30 ( 3.14 to 1.39) Frequency 22.5 (9.8) 21.9 (11.9) 0.60 ( 0.04 to 1.21).07 Air Conduction 0.5 khz 56.1 (13.2) 29.3 (14.9) (25.64 to 28.02) khz 52.9 (14.0) 27.0 (14.8) (24.69 to 27.07) khz 47.8 (15.0) 28.8 (15.5) (17.91 to 20.14) khz 49.6 (20.8) 37.3 (21.2) (10.95 to 13.57) Frequency 51.6 (13.5) 30.6 (14.4) (19.96 to 22.04).001 Air-Bone Gap 0.5 khz 40.2 (12.0) 12.3 (10.9) (26.76 to 28.96) khz 34.4 (10.8) 9.4 (9.5) (24.00 to 26.02) khz 18.1 (11.2) 3.8 (7.0) (13.34 to 15.20) khz 23.8 (13.3) 9.2 (10.5) (13.40 to 15.66) Frequency 29.1 (9.3) 8.7 (7.4) (19.60 to 21.23).001 curacy of the models was estimated on the basis of their reliability using the Hosmer-Lemeshow goodness-of-fit test. 16 The ability of a model to discriminate between patients was estimated as the area under the receiver operating characteristic curve of the model. 17 The receiver operating characteristic curve area is a suitable variable to summarize the discriminative or predictive value and can range from 0.5 (no discrimination, like the flip of a coin) to 1.0 (perfect discrimination). In addition, we calculated absolute risks of a postoperative ABG of 10 db or less or a postoperative gain in AC exceeding 20 db across combinations of independent predictors. MISSING VALUES Information was available for 82.4% of the predictor variables and for 71.3% of the outcome variables. Data are seldom missing at random; it has been shown that removal of participants with a missing value for 1 of the predictors studied (complete case analysis) commonly leads to biased results and certainly to loss of power. 18,19 To decrease bias and to increase statistical efficiency, it is better to impute missing data than to perform a complete case analysis. 20,21 Accordingly, we imputed the missing data for each trial using multiple imputation (SPSS 17.0 for Windows; SPSS, Inc, Chicago, Illinois). Such imputation is based on the correlation between each variable with missing values and all other variables, as estimated for the set of participants with complete data. RESULTS Nine hundred thirty-nine patients with otosclerosis underwent primary stapes surgery between January 1, 1982, and March 1, Preoperative and postoperative audiometric data were available for 666 of them. The mean (SD) age of the total cohort was 41.4 (11.5) years and ranged from 9 to 74 years, with 36.6% male and 63.4% female. The operated ears comprised 49.3% left ears and 50.7% right ears. Introduction of the potassium titanyl phosphate laser at our institution in 2007 resulted in 1.5% of the surgical procedures being performed using this modality. A total of 85.2% of patients were operated on by one of us (R.A.T.), and 14.8% of patients were operated on by another of us (W.G.). Otosclerosis was bilateral in 31.1% of patients. Table 1 summarizes characteristics of the total cohort (N=939) and of patients with preoperative and postoperative audiometric data available (n=666) (ie, those without missing values). No substantial differences were observed between the groups. PREOPERATIVE AND POSTOPERATIVE AUDIOMETRIC EVALUATION AMONG 666 PATIENTS Preoperative and postoperative audiometric data (n=666) are summarized in Table 2. The mean preoperative ABG, AC, and BC improved by db (95% confidence interval [CI], to db), db (19.96 to db), and 0.60 db ( 0.04 to 1.21), respectively, after surgery. PROGNOSTIC MODEL BASED ON A POSTOPERATIVE ABG OF 10 db OR LESS Among 666 patients, 72.1% (n=480) had a postoperative ABG of 10 db or less, and 93.8% (n=625) had a postoperative ABG of 20 db or less. Age at surgery and preoperative ABG and AC were univariate and independent predictors of a postoperative ABG of 10 db or less. The prognostic model showed a good fit (P=.97, Hosmer- Lemeshow goodness-of-fit test), and the area under the receiver operating characteristic curve was 0.62 (95% CI, ). Table 3 gives absolute risks of a postoperative ABG of 10 db or less with certain combinations of independent predictors. A patient older than 40 years with a preoperative ABG exceeding 30 db has a 67.9% chance of achieving a postoperative ABG of 10 db or less, whereas a patient 40 years or younger with an ABG exceeding 30 db has a 60.6% chance. Analysis of the data using multiple imputation showed similar results. 782

4 Table 3. Absolute Risks Based on Postoperative 4-Frequency Air-Bone Gap (ABG) of 10 db or Less Among 666 Patients Without Missing Values % Table 4. Absolute Risks Based on Postoperative 4-Frequency Gain in Air Conduction (AC) Exceeding 20 db Among 666 Patients Without Missing Values % Combination of Independent Predictors Postoperative ABG 10 db (n=480) Postoperative ABG 10 db (n=186) Combination of Independent Predictors and and and and Postoperative Gain in AC 20 db (n=372) Postoperative Gain in AC 20 db (n=294) PROGNOSTIC MODEL BASED ON A POSTOPERATIVE GAIN IN AC EXCEEDING 20 db Among 666 patients, 55.9% (n=372) had a postoperative gain in AC exceeding 20 db, and 44.1% (n=294) had a postoperative gain in AC of 20 db or less. Univariate predictors of a postoperative gain in AC exceeding 20 db were preoperative ABG, AC, and BC. Independent predictors of a postoperative gain in AC exceeding 20 db were preoperative ABG and AC. The prognostic model showed a good fit (P=.34, Hosmer-Lemeshow goodnessof-fit test), and the area under the receiver operating characteristic curve was 0.80 (95% CI, ). Table 4 gives absolute risks of a postoperative gain in AC exceeding 20 db with certain combinations of independent predictors. A patient with a preoperative AC exceeding 50 db and a preoperative ABG exceeding 30 db has an 86.2% chance of achieving a postoperative gain in AC exceeding 20 db, whereas a patient with a preoperative AC of 50 db or less and a preoperative ABG exceeding 30 db has a 69.0% chance. Analysis of the data using multiple imputation showed similar results. COMMENT Three months after primary stapes surgery for otosclerosis, 72.1% of patients had a postoperative ABG of 10 db or less, and 93.8% of patients had a postoperative ABG of 20 db or less. Age at surgery and preoperative ABG were independent prognostic determinants of achieving a postoperative ABG of 10 db or less. Preoperative ABG and AC were independent prognostic determinants of achieving a postoperative gain in AC exceeding 20 db. Our results are in agreement with Kisilevsky et al, 6 who reported a mean postoperative ABG of 10 db or less in 75.2% (n=861) of their patients using the mean thresholds at 0.5, 1, 2, and 4 khz. However, the authors reported a mean follow-up period of 16.4 months (range, months) compared with a mean follow-up period of 3 months in our series. Our results are not in agreement with Vincent et al, 7 who reported a mean postoperative 4-frequency ABG of 10 db or less in 95.5% (n=1838) of their patients using the mean thresholds at 0.5, 1, 2, and 4 khz, with a follow-up period of less than 1 year (range, 3-11 months). Abbreviation: ABG, air-bone gap. The authors also reported long-term audiometric results per year and showed that the percentage of patients with a postoperative ABG of 10 db or less was consistently around 96%. Among patients in their study, 95.6% (n=800) had a mean postoperative ABG of 10 db or less at a mean follow-up period of 1 year (range, months). There is large variability in surgical outcomes. Possible explanations are surgical technique and surgical experience. 22,23 For example, Vincent et al 7 interpositioned a vein graft 24 between the fenestration and the prosthesis among patients in their series, representing one of the largest samples of patients with otosclerosis in the literature. Our prognostic model is in agreement with results by Marchese et al, 13 who also found that age and preoperative ABG were the strongest predictors of postoperative success. However, they evaluated prognostic factors affecting gain in ABG, which is different from actual postoperative ABG. These outcome measurements are related, but a gain in ABG does not necessarily result in a postoperative 4-frequency ABG of 10 db or less. Our finding that a smaller preoperative ABG increases the chance of better postoperative ABG closure is also congruent with findings by Ueda et al. 10 In 2003, Welling et al 25 published an article about predictive factors in stapes surgery, but they focused on pediatric patients only. Our results (Table 3) show that older patients ( 40 years) have a slight advantage over younger patients ( 40 years) in achieving a postoperative ABG of 10 db or less. This finding is in contrast to Marchese et al, 13 who reported that older patients ( 50 years) had a lower probability of a good functional outcome after stapedotomy. There is a clear correlation between patient age and degree of advanced otosclerosis. However, based on the present results, we are unable to explain the operative mechanism between patient age and pathologic otosclerosis. The major strength of our study is that we were able to identify independent prognostic factors affecting postoperative ABG of 10 db or less or postoperative AC gain exceeding 20 db. These prognostic factors were combined and translated into absolute risks for patients. The cutoff points in Tables 3 and 4 (age and preoperative ABG and AC) were used for applicability of absolute risk tables in 783

5 our clinic during preclinical assessment. The mean age of the study cohort was 41.4 years; therefore, we used 40 years as a cutoff for age at surgery. A preoperative ABG of 30 db was used because it renders a patient eligible for a hearing device. In other words, a patient with primary otosclerosis and an ABG of 30 db could choose between primary stapes surgery or a hearing device. The mean preoperative AC was 52 db; therefore, we used 50 db as a cutoff for preoperative AC. The actual prognostic model (multivariate logistic regression analysis) was unaffected by cutoff points and only shows which factors significantly affect the chances of achieving a postoperative ABG of 10 db or less and a postoperative gain in AC exceeding 20 db. To our knowledge, this is the first publication focusing on such a prognostic model and its related absolute risks in primary stapes surgery. Some possible limitations should also be discussed. First, about 18% of the predicting variables and 29% of the outcome variables were missing, which might lead to bias and loss of power. 18,19 However, the results after using multiple imputation were in agreement with the complete case analyses. Second, because other surgeons may operate on slightly different patients or use other techniques and because surgical skills might differ among surgeons, our results may not be directly applicable to other clinics and surgeons. Third, the method of reporting audiometry results in stapes surgery could affect the final results. 9 We expect that the results of this study will improve understanding of prognostic characteristics among patients with otosclerosis who undergo primary stapes surgery. In conclusion, following primary stapes surgery, postoperative ABG of 10 db or less and postoperative gain in AC exceeding 20 db may be predicted with accuracies of 62.1% and 80.1%, respectively. The chance of achieving a postoperative ABG of 10 db or less was 1.3 times higher in patients older than 40 years with a preoperative ABG of 30 db or less compared with patients 40 years or younger with a preoperative ABG exceeding 30 db. The chance of achieving a postoperative gain in AC exceeding 20 db was 2.5 times higher in patients having a preoperative AC exceeding 50 db and a preoperative ABG exceeding 30 db compared with patients having a preoperative AC of 50 db or less and a preoperative ABG of 30 db or less. Clinicians can use these factors to inform patients more explicitly about expected postoperative audiometric results. Submitted for Publication: December 9, 2010; final revision received April 10, 2011; accepted April 19, Published Online: July 18, doi: /archoto Correspondence: Arnold J. N. Bittermann, MD, Department of Otorhinolaryngology Head and Neck Surgery, University Medical Center, Heidelberglaan 100, Room G02.129, 3584 CX Utrecht, the Netherlands (ENT-Research@umcutrecht.nl). Author Contributions: Drs Bittermann, Rovers, Tange, and Grolman had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Bittermann, Rovers, Tange, and Grolman. Acquisition of data: Bittermann, Tange, Vincent, Dreschler, and Grolman. Analysis and interpretation of data: Bittermann, Rovers, Tange, and Grolman. Drafting of the manuscript: Bittermann. Critical revision of the manuscript for important intellectual content: Bittermann, Rovers, Tange, Vincent, Dreschler, and Grolman. Statistical analysis: Bittermann and Rovers. Administrative, technical, and material support: Bittermann and Dreschler. Study supervision: Rovers, Tange, Vincent, and Grolman. Financial Disclosure: None reported. REFERENCES 1. Menger DJ, Tange RA. The aetiology of otosclerosis: a review of the literature. Clin Otolaryngol Allied Sci. 2003;28(2): Schrauwen I, Van Camp G. The etiology of otosclerosis: a combination of genes and environment. Laryngoscope. 2010;120(6): Declau F, Van Spaendonck M, Timmermans JP, et al. Prevalence of otosclerosis in an unselected series of temporal bones. Otol Neurotol. 2001;22(5): Kisilevsky V, Bailie NA, Halik JJ. Bilateral hearing results of 751 unilateral stapedotomies evaluated with the Glasgow benefit plot. J Laryngol Otol. 2010;124 (5): de Bruijn AJ, Tange RA, Dreschler WA. Efficacy of evaluation of audiometric results after stapes surgery in otosclerosis, II. Otolaryngol Head Neck Surg. 2001; 124(1): Kisilevsky VE, Dutt SN, Bailie NA, Halik JJ. Hearing results of 1145 stapedotomies evaluated with Amsterdam hearing evaluation plots. J Laryngol Otol. 2009; 123(7): Vincent R, Sperling NM, Oates J, Jindal M. Surgical findings and long-term hearing results in 3,050 stapedotomies for primary otosclerosis: a prospective study with the otology-neurotology database. Otol Neurotol. 2006;27(8)(suppl 2):S25-S Lippy WH, Burkey JM, Schuring AG, Rizer FM. Stapedectomy in patients with small air-bone gaps. Laryngoscope. 1997;107(7): de Bruijn AJ, Tange RA, Dreschler WA. Efficacy of evaluation of audiometric results after stapes surgery in otosclerosis. Otolaryngol Head Neck Surg. 2001; 124(1): Ueda H, Miyazawa T, Asahi K, Yanagita N. Factors affecting hearing results after stapes surgery. J Laryngol Otol. 1999;113(5): Gerard JM, Serry P, Gersdorff MC. Outcome and lack of prognostic factors in stapes surgery. Otol Neurotol. 2008;29(3): Moons KG, Royston P, Vergouwe Y, Grobbee DE, Altman DG. Prognosis and prognostic research: what, why, and how? BMJ. 2009;338:b Accessed May 4, Marchese MR, Conti G, Cianfrone F, Scorpecci A, Fetoni AR, Paludetti G. Predictive role of audiological and clinical features for functional results after stapedotomy. Audiol Neurootol. 2009;14(5): Marquet J. Otosclerosis: small hole technique. J Laryngol Otol Suppl. 1983;8: American Academy of Otolaryngology Head and Neck Surgery Foundation, Inc. Committee on Hearing and Equilibrium guidelines for the evaluation of results of treatment of conductive hearing loss. Otolaryngol Head Neck Surg. 1995;113 (3): Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons Inc; 1989: Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143(1): Greenland S, Finkle WD. A critical look at methods for handling missing covariates in epidemiologic regression analyses. Am J Epidemiol. 1995;142(12): Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15(4): van der Heijden GJ, Donders AR, Stijnen T, Moons KG. Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: a clinical example. J Clin Epidemiol. 2006; 59(10): Donders AR, van der Heijden GJ, Stijnen T, Moons KG. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol. 2006;59(10): Yung MW, Oates J, Vowler SL. The learning curve in stapes surgery and its implication to training. Laryngoscope. 2006;116(1): Hughes GB. The learning curve in stapes surgery. Laryngoscope. 1991;101(12, pt 1): Causse JB, Causse JR, Parahy C. Stapedotomy technique and results. Am J Otol. 1985;6(1): Welling DB, Merrell JA, Merz M, Dodson EE. Predictive factors in pediatric stapedectomy. Laryngoscope. 2003;113(9):